Adding dual kinds of nanoparticle fillers with higher thermal conductivity (TC) into matrix materials provides versatile and effective means for enhancing the TC of polymer materials. In this study, the effective thermal conductivity (ETC) of composite materials with heterogeneous-filler-nanoparticles is investigated based on an extensive FEM numerical study, in terms of the TC ratios between the filler particles and the matrix material (kappa(1) and kappa(2)), and the volume fractions (VFs) of each filler particle (phi(1) and phi(2)). The results indicate that the ETC of composite material significantly depends on TC ratio between two fillers (kappa(1)/kappa(2)). In addition, there exists a maximum ETC according to each sum of TC ratio (kappa(1) + kappa(2)). The asymmetric behavior of ETC in terms of the TC ratio between two fillers becomes symmetric when two fillers has the same VF. Based on the numerical results for a wide range of geometric parameters, a generalized correlation for predicting ETC is proposed as a function of four non-dimensional parameters: phi(1), phi(2), kappa(1), and kappa(2). The correlation is valid for the entire practical range of parameters (0 < phi(1) or phi(2) < 0.4; 1 < kappa(1) or kappa(2) < 10(4)), and can be widely utilized for predicting the TC of nanoparticle-added composite materials. (C) 2015 Elsevier Ltd. All rights reserved.